StringIsEmpty with Refaster test fails

[Uriziel47]

Good Morning,

I just wanted to get started on openrewrite. So I decided to check this Repository first and get my hands dirty.
Unfortunately I got stucked in the first try using Refaster.

So after writing the Refaster recipe of StringIsEmpty. I didn't get the testcase, StringIsEmpty::standardizeStringIsEmpty to pass. Either the case of string.equals("") or string.length == 0 case is wrong.

So after inspecting the generated class. I realized, that the first @BeforeTemplate annotation decides which visitor method will be used for all @BeforeTemplate annotated methods.

In case string.equals("") the visitor method visitMethodInvocation is been used. And in case of string.length == 0 the visitor method visitBinary is been used.

So I am just curious, if I misunderstood the documentation, used the wrong version, or is this a bug, which I did not find in the issues.

For clarification I created a branch in my learning repository: link to repo.
In there the discussed variants with the testcases.

• StringIsEmptyCheckEqualsFirst
• StringIsEmptyCheckLengthFirst
• StringIsEmptyWithSubclasses (splitted equals and length into refaster subclasses)
• StringisEmptyVisitorRecipe (The recipe I thought it would be generated)

Thanks for your answers in advance.

[timtebeek]

Hi @Uriziel47 ; thanks for the detailed report! This is indeed a bug, that we'll track against the refaster template processor:

• Visit multiple elements for diverse before templates openrewrite/rewrite-templating#117

Let's continue the conversation there! Any fix & release would also resolve that issue here, without a need for changes in this project beyond bumping the version.

[Uriziel47]

Hi @timtebeek, thank you very much for the fast feedback and fix!

[timtebeek]

You're welcome; your report made it easy to pick up. You'll have noticed I also released a new version of rewrite-templating and bumped the versions here. We'll do a full new release this Wednesday. Thanks again!